turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- proc glmselect example

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

12-16-2015 07:45 AM

I'm trying to understand the proc glmselect with simple example. However I could not find the example which can be easily understandable.

May I request someone to provide or explain me the basic features of proc glmselect in simple terms?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Babloo

12-16-2015 08:26 AM

Try these papers:

Paper 259-2009

Applications of the GLMSELECT Procedure for Megamodel Selection

Robert A. Cohen, SAS Institute Inc., Cary, NC

https://support.sas.com/resources/papers/proceedings09/259-2009.pdf

Paper 207-31 (2006)

Introducing the GLMSELECT PROCEDURE for Model Selection

Robert A. Cohen, SAS Institute Inc. Cary, NC

http://www2.sas.com/proceedings/sugi31/207-31.pdf

and a recent (2014) paper/presentation as well.

Model Selection for Linear Models with SAS/STAT Software

Funda Gunes

SAS Institute Inc.

http://www.amstat.org/meetings/csp/2014/onlineprogram/handouts/T3-Handouts.pdf

Koen

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Babloo

12-16-2015 08:37 AM

The basic idea is that you have a response variable that you want to model, and a candidate list of hundreds of possible explanatory variables. You are not sure which explanatory variables are important and which are not. That is, which variables are good for predicting the response?

The GLMSELECT procedure enables you to throw hundreds of candidate variables into a MODEL statement. It supports running various algorithms that try to produce a parsimonious model based on those candidate variables. In theory, the data themselves choose the variables that are important, rather than the analyst.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Rick_SAS

12-17-2015 05:12 AM

Thank you for the explanation. May I request you to give the simple example?

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Babloo

12-17-2015 06:19 AM

Certainly. The "Getting Started" example in the PROC GLMSELECT documentation is a simple example. It provides a discussion of syntax and a discussion of how to interpret the output.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Babloo

12-16-2015 08:45 AM

What exactly you want to do using **proc glmselect ...**

SAS/STAT(R) 9.3 User's Guide: The GLMSELECT Procedure Examples